【作者】 陈黎明；

【导师】 苏锡坤；

【作者基本信息】 华南理工大学 ， 管理科学与工程， 2017， 硕士

【摘要】 本文利用日内高频数据来构造一个夏普指数序列,同时利用VaR值对风险进行调整,得到基于VaR调整的高频夏普指数序列,基于该指数序列,提出一种资产选择和组合策略的有效方法。将带总头寸限制的M-V问题转化到带惩罚的回归问题上,进而可以有效地求解,我们根据此思路,将新构造的基于VaR调整的高频夏普指数放入回归问题的框架中进行资产选择,然后确定组合权重得到最优资产组合。资产选择与最优组合权重的设置是构建投资组合的两个关键步骤。在实证中,利用日内高频数据构建一个基于VaR调整的夏普指数序列来进行资产选择,同时考虑多种组合策略。以沪深300市场数据进行样本外实证分析,不论市场处于下行还是上行行情,基于VaR调整的高频夏普指数选股方法构建的组合策略与其他的选股方法构建的组合策略进行比较,如动量选股方法和未进行VaR调整的高频夏普指数选股方法,看是否能够取得更好的超额收益。基于VaR调整的夏普指数序列来进行资产选择方法所取得的收益,我们选择了未进行VaR调整的夏普指数序列的资产选择选择方法以及动量选股作为比较对象,在实证中发现,前者取得的收益明显比后两者更加稳定,说明基于VaR调整之后的夏普指数序列,一定程度上消除了波动对收益的影响,使选择的资产在后续的表现中更加稳定,达到了我们想要的效果。更多还原

【Abstract】 This paper uses the intraday high frequency data to construct a VaR adjusted sharp index series.Based on this exponential sequence,we propose an effective method of asset selection and combination strategy.We can transfer the mean-variance problem with the total position constraint into a penalized regression problem,which can be solved efficiently,according to this method,the newly constructed high frequency sharp index based on VaR adjustment is put into the framework of regression problem for asset selection,and we look for proper weights of selected assets to get the optimal portfolio.Asset selection and optimal combination weighting are two key steps in building a portfolio.In the empirical study,we use the daily high frequency data to construct a Sharpe index series based on VaR adjustment to carry out asset selection,which is took into account various combinations of strategies.Based on the empirical analysis of the CSI 300 market data,whether the market is down or up,compare the combination strategy constructed through high frequency VaR adjusted sharp index method with that by other stock selection methods,such as momentum stock selection method and stock selection method based on high frequency sharp index,we want to know if the new method can get better returns.To illustrate the effectiveness of the asset selection method based on high frequency sharp index,we choose momentum stock selection method and stock selection method based on high frequency sharp index as the comparative objects,we can find in the empirical research,the former gains significantly more stable than the last two,the asset selection method based on high frequency sharp index can eliminate the influence of the volatility of the assets a certain extent,and the asset selection also get more stable returns in subsequent performance,we achieved the desired effect.更多还原